There are two types of programmable sequencer: a process incrementing type in which logic operation is carried out according to the sequence program each time an external signal is entered; and a cyclic type in which logic operation is executed according to the sequence program in synchronism with a timing signal irrespective of the arrival of an input signal. The process incrementing type is used in applications where the number of input and output points is relatively small and therefore such functions as interrupt processings are not required. In the cyclic type the time required for execution of one cycle of the sequence program depends on the length of the sequence program. In recent years, sequence programs exceeding 100 milliseconds are available. With such long sequence programs, each step of the sequence program can only be executed once for every 100 milliseconds. When a signal indicating the occurrence of abnormal condition is entered, however, there may be cases where a processing required by the emergency situation cannot be delayed until the next scanning. To deal with such a situation an interrupt processing function is needed.
As the number of input and output points increases, the number of input signals that require an interrupt processing may reach as many as several tens of signal points. In such cases, however, if identifying the signal source that produced an interrupt takes too much time, the merit of using the interrupt request signal may be offset resulting in a delayed response.
Among the sequence controllers with an interrupt processing function, the Japanese Utility Model Laid-Open No. 198645/1982 and the Japanese Patent Laid-Open No. 145329/1987 may be cited.